Design method for an anthropomorphic hand able to gesture and grasp

TL;DR

This paper introduces a numerical design method for creating anthropomorphic robotic hands that can gesture and grasp, focusing on thumb placement to enhance dexterity and grasping ability.

Contribution

It presents a novel numerical approach for finger and thumb placement in human-like robotic hands, incorporating empirical testing and workspace analysis.

Findings

Validated through simulation for balanced and functional hand design

Provides a numerical tool for easy computation of finger and thumb geometries

Improves hand dexterity and grasping capabilities by optimizing thumb placement

Abstract

This paper presents a numerical method to conceive and design the kinematic model of an anthropomorphic robotic hand used for gesturing and grasping. In literature, there are few numerical methods for the finger placement of human-inspired robotic hands. In particular, there are no numerical methods, for the thumb placement, that aim to improve the hand dexterity and grasping capabilities by keeping the hand design close to the human one. While existing models are usually the result of successive parameter adjustments, the proposed method determines the fingers placements by mean of empirical tests. Moreover, a surgery test and the workspace analysis of the whole hand are used to find the best thumb position and orientation according to the hand kinematics and structure. The result is validated through simulation where it is checked that the hand looks well balanced and that it meets our constraints and needs. The presented method provides a numerical tool which allows the easy computation of finger and thumb geometries and base placements for a human-like dexterous robotic hand.